To evaluate whether the hydroxylated metabolites of quinidine (Q) and hydroquinidine (HQ): hydroxy-3S-quinidine (OH-Q) and hydroxy-3S-hydroquinidine (OH-HQ), exert electrophysiologic effects and participate in the therapeutic action of the parent drugs, we examined and compared the effects of the metabolites and the parent drugs on the electrical activity of guinea pig ventricular cells recorded by standard microelectrode technique. In addition, we investigated the potential arrhythmogenic properties of these compounds in rabbit Purkinje fibers in low K+ (2.7 mM) Tyrode's solution. The concentration [C]-, frequency-, and voltage-dependent effects of the drugs were investigated. Maximum upstroke velocity of phase 0 (Vmax) was [C]-dependently depressed by both OH-Q and OH-HQ but at a lesser degree than with Q and HQ, respectively: at the [C] of 50 microM, Vmax depression attained 26.7 +/- 2.6% with OH-Q versus 45.9 +/- 1.6% with Q and 32.3 +/- 1.9% with OH-HQ versus 54.6 +/- 1.4% with HQ. This effect was frequency and voltage dependent without significant differences between the four compounds. In the presence of equipotent [C], recovery kinetics of Vmax was significantly slower with metabolites than with respective parent drugs. In contrast, the effects of metabolites on action potential duration at 90% of repolarization (APD90) and effective refractory period (ERP) differed from those observed with parent drugs. With metabolites, APD90 and ERP were increased in a [C]-dependent manner, whereas the Q- and HQ-induced lengthening in APD90 and ERP was observed only at low concentration and low frequency. In addition, the OH-Q- and OH-HQ-induced APD90 lengthening was not altered by increasing pacing rate. In rabbit Purkinje fibers, increase in cycle length and prolonged exposure to either metabolites or parent drug caused early afterdepolarizations (EADs) and triggered activity. With all drugs tested, EADs arose more frequently at the plateau level than at the final repolarization of AP, but the incidence of EADs appeared to be much lower with metabolites than with parent drugs. The present results demonstrate that OH-Q and OH-HQ exert qualitatively similar but quantitatively less potent depressant effects on Vmax than Q and HQ, respectively, but differ in the lengthening effect on APD. We suggest that metabolites may participate in class I antiarrhythmic action of their respective parent drug and contribute to their arrhythmogenicity.